K M Vara Prasad et al.

, International Journal of Research in Engineering, IT and Social Sciences, ISSN 2250-0588, Impact Factor:
6.565, Volume 13 Issue 05, May 2023, Page 238-242

Detection of Bike Riders without Helmet and

Triple Rides using Deep Learning
Dr. K M Vara Prasad1, Y. Udaykiran2, U. Chandra Sekhar3, P. Basha4, Y. Rajesh5, T. Srivatsava6

1 Professor, 2Student, 3Student, 4Student, 5Student, 6Student

1,2,3,4,5,6
Department of Computer Science and Engineering/ Narayana Engineering College/ Nellore/ Andhra
Pradesh/ India

ABSTRACT

Violations in traffic laws are very common in a highly populated country like India. The accidents associated with these violations cause a
huge loss to life and property. Since utilization of bikes is high, mishaps associated with bikes are additionally high contrasted with
different vehicles. One of the main causes of these is not using motorcycle helmets. So we propose an approach called “TRAFFIC
VIOLATION PROCTORING SYSTEM:HELMET AND TRIPLE RIDING DETECTION” using deep learning which automatically
sends challan or send an SMS for individuals in case of identification of bicycle riders without headgear and who are triple riding utilizing
surveillance videos in real-time. The proposed approach initially recognizes motorcycle riders utilizing background subtraction and object
segmentation. At that point we utilize object classifier to classify violators. Since wearing helmet is critical while driving, our main aim is
to decrease the danger of injuries in case of accident. By detecting the motorcyclists without helmets, triple riding or other violations we
can therefore increase their safety while on road. Hence by automating we reduce the workload on the traffic control team and will be
able to share the evidence with the team efficiently to impose fines on violators.

Keywords : YOLO Algorithm, COCO Dataset, CNN.

I. INTRODUCTION

Motorcycle accidents have been rapidly growing throughout the years in many countries. Due to various social and
economic factors, this type of vehicle is becoming increasingly popular. The helmet is the main safety equipment of
motorcyclists, however many drivers do not use it. The main goal of helmet is to protect the drivers head in case of accident. In
case of accident, if the motorcyclist does not use can be fatal. This project aims to propose a system for detection of
motorcyclist without helmet. Automation of this process is highly desirable for reliable and robust monitoring of these
violations as well as it also significantly reduces the amount of human resources needed. Also, many countries are adopting
systems involving surveillance cameras at public places. So, the solution for detecting violators using the existing infrastructure
is also cost-effective. This work proposes a system for location of individual or different riders taking a trip on bikes with no
helmets.

II. WORKING OF A PROJECT

This project identifies whether a man or woman wearing a helmet while travelling on the two wheeler or not. Also,
identifies the triple rides in the motor cycles and captures the images and send the mail to the authenticated vehicle owner. So,
that particular person can knew that the challan is issued.

http://indusedu.org Page 238

This work is licensed under a Creative Commons Attribution 4.0 International License
K M Vara Prasad et al., International Journal of Research in Engineering, IT and Social Sciences, ISSN 2250-0588, Impact Factor:
6.565, Volume 13 Issue 05, May 2023, Page 238-242

III. DATASET COLLECTION AND PREPROCESSING

The first step in data collection is to determine the scope of the dataset. This includes deciding on the number of images and the
variety of scenes needed for your object detection model. You can collect your own images using a camera system, either by taking pictures of
vehicles on the road or by setting up a controlled environment. Another option is to use pre-existing datasets that contain images of vehicles with
helmets and number plates. For this project, the dataset should include images of vehicles with visible helmets and number plates, covering
different lighting conditions, angles, and distances to ensure that the model can detect helmets and number plates accurately in real-world
scenarios.

IV. YOLO MODEL TRAINING

Model training is a crucial step in creating an accurate and reliable object detection system for helmet detection with number plate
detection and challan generation. The first step in model training is selecting an appropriate object detection model, such as the YOLO algorithm,
which is commonly used for its speed and accuracy. Next, the dataset is prepared for training by splitting it into training, validation, and testing sets.
The training set is used to train the model, while the validation set is used to tune the model's hyperparameters, such as learning rate, batch size, and
number of epochs are adjusted during training to improve the model's performance.

V. E-CHALLAN FOR TRIPLE RIDING

After detection of the license plate number, the system generates E-challan to the offenders by accessing the database for their
details. It sends the message through E-mail of the offender

VI. RESULTS & OUTPUT

SAMPLE DATABASE

http://indusedu.org Page 239

This work is licensed under a Creative Commons Attribution 4.0 International License
K M Vara Prasad et al., International Journal of Research in Engineering, IT and Social Sciences, ISSN 2250-0588, Impact Factor:
6.565, Volume 13 Issue 05, May 2023, Page 238-242

HELMET AND NUMBER PLATE RECOGNITION

OUTPUT FOR TRIPLE RIDING DETECTION

http://indusedu.org Page 240

This work is licensed under a Creative Commons Attribution 4.0 International License
K M Vara Prasad et al., International Journal of Research in Engineering, IT and Social Sciences, ISSN 2250-0588, Impact Factor:
6.565, Volume 13 Issue 05, May 2023, Page 238-242

TICKET FOR NOT WEARING HELMET

TICKET INFORMATION

http://indusedu.org Page 241

This work is licensed under a Creative Commons Attribution 4.0 International License
K M Vara Prasad et al., International Journal of Research in Engineering, IT and Social Sciences, ISSN 2250-0588, Impact Factor:
6.565, Volume 13 Issue 05, May 2023, Page 238-242

VII. FUTURE ENHANCEMENT

In the future, the challan generation module could be integrated with automated payment systems to enable violators to pay their fines
online, making the payment process more convenient and efficient. Implementation in other locations: The system could be deployed in other
locations to promote road safety and reduce traffic violations on a larger scale. Integration with other traffic management systems: The system could
be integrated with other traffic management systems, such as traffic signal control systems or speed detection systems, to create a more
comprehensive approach to road safety. The accuracy of the object detection model could be improved by using more advanced techniques, such as
deep learning or neural networks. The user interface of the system could be improved to make it more user-friendly and intuitive, making it easier
for users to interact with the system.

VIII. CONCLUSION
We have applied our approach to video data from diverse road environments, which included adverse factors such as occlusion,
differences in camera angle, an imbalanced number of coded classes, as well as differing rider numbers per motorcycle and varying
traffific densities. For future research, we intend to enhance the HELMET dataset by incorporating scenes with more diverse traffific
infrastructure, e.g. crossroads, to ensure more robust application of the approach. Also, more training data that contain sparked
motorcycles will be acquired and used for training so that these objects will not be detected as false positive more advanced techniques,
such as deep learning or neural networks. The user interface of the system could be improved to make it more user-friendly and intuitive,
making it easier for users to interact with the system.

ACKNOWLEDGEMENT

I would like to thank my teachers for providing invaluable guidance and support throughout the research process. Their knowledge
and expertise were instrumental in shaping the direction of this study. I would also like to extend my appreciation to the reference base papers
that served as a foundation for this research. Without their thorough analysis and insights, this study would not have been possible. In
addition, I would like to thank my colleagues and friends who provided valuable feedback and support throughout the writing process. Their
critiques and encouragement were invaluable.Thank you all for your support and encouragement throughout this project.

REFERENCES

[1] Lijun Wang, Yunyu Cao, Song Wang, “Investigation Into Recognition Algorithm of Helmet Violation Based on YOLOv5-CBAM-DCN”, 2022
IEEE Access ( Volume: 10), 2022.

[2] BAINING ZHAO , HAIJUAN LAN , ZHEWEN NIU , HUILING ZHU, “Detection and Location of Safety Protective Wear in Power Substation
Operation Using Wear-Enhanced YOLOv3 Algorithm” 2021 IEEE Access, 2021.

[3] Linu Shine ·Jiji C. V., ” Automated detection of helmet on motorcyclists from traffic surveillance videos: a comparative analysis using hand-
crafted features and CNN” 2020 Springer Science + Business Media, 2020.

[4] Felix Wilhelm Sieberta, HanheLinb, “Detecting motorcycle helmet use with deep learning”, 2019 https://doi.org/10.1016/j.aap.2019.105319
[5] HANHE LIN , JEREMIAH D. DENG , (Member, IEEE), DEIKE ALBERS , AND FELIX WILHELM SIEBER, “Helmet Use Detection of
Tracked Motorcycles Using CNN-Based Multi- Task Learning”, 2020 IEEE Access, 2020.

[6] KUN HAN AND XIANGDONG ZENG, “Deep Learning-Based Workers Safety Helmet Wearing Detection on Construction Sites Using Multi-
Scale Features” 2021 IEEE Access, 2021.

[7] Ahatsham Hayat and Fernando Morgado-Dias, “ Deep Learning-Based Automatic Safety Helmet Detection System for Construction Safety”
2022 Applied sciences 2022.

[8] R. Vandana , M. Hari Krishna, Ch. Anusha, “Helmet OR TRIPLE RIDIND DETECTION USING DEEP LEARNING” 2022 European Journal
of Molecular & Clinical Medicine 2022.

[9] Lokesh Allamki, Manjunath Panchakshari, Ashish Sateesha, K S Pratheek, “Helmet Detection using Machine Learning and Automatic License
Plate Recognition” 2019 International Research Journal of Engineering and Technology (IRJET) 2019.

[10] Jie Li, Huanming Liu, Tianzheng Wang, and Min Jiang, “Safety Helmet Wearing Detection Based on Image Processing and Machine Learning”
2017 Ninth International Conference on Advanced Computational Intelligence 2017.

http://indusedu.org Page 242

This work is licensed under a Creative Commons Attribution 4.0 International License